Anti-reverse power transmission state detection method and system for power distribution network maintenance

Abstract

The invention provides an anti-reverse power transmission state detection method and system for power distribution network maintenance, and the method comprises the steps: extracting a voltage amplitude feature, a current amplitude feature and a phase difference feature corresponding to each harmonic wave according to a preset harmonic wave order set, and extracting a ratio feature of first frequency band energy to second frequency band energy, the phase change rate between adjacent frames is calculated to form a node feature vector, and the frequency of the first frequency band is greater than that of the second frequency band; discretizing each component of the node feature vector according to a preset quantization boundary set, and generating a fingerprint abstract corresponding to the node through a fixed mapping relation and a Hash algorithm; and performing consistency screening on the nodes according to the fingerprint abstract, if the consistency is inconsistent, performing fine-grained comparison on the nodes, and outputting a final detection result according to a comparison result, the comparison result comprising a maximum deviation value and a consistency degree. According to the invention, the accuracy of reverse power transmission prevention state detection can be improved.

Description

Technical Field

[0001] This invention relates to the field of reverse power transmission status detection technology, and in particular to a method and system for reverse power transmission status detection used in distribution network maintenance. Background Technology

[0002] With the modernization of power systems, distribution networks play a crucial role in ensuring power supply security and enhancing grid flexibility. In recent years, in particular, the large-scale integration of distributed power sources and renewable energy sources (such as solar and wind power) has made power system operation more complex. To ensure the safety of distribution networks during maintenance, prevent backfeeding, and protect equipment and personnel, monitoring and protection against backfeeding risks during maintenance has become a critical issue that urgently needs to be addressed in the distribution network sector.

[0003] Currently, the backfeeding problem in distribution network maintenance mainly refers to the reverse current flow within the power grid caused by the erroneous connection of external or local power sources during the maintenance of power equipment. This can even affect equipment and systems in other work areas, causing equipment damage or personal injury. To address this, many traditional backfeeding detection methods have emerged; however, existing technologies have many shortcomings in handling backfeeding risks in complex distribution network environments.

[0004] However, some existing technologies rely on reverse power or simple current direction detection, making it difficult to accurately determine the existence of backfeeding risks. Especially in complex distribution environments, such as those with distributed generation or multiple feeders, existing methods cannot accurately identify hidden backfeeding paths. Furthermore, existing technologies often fail to achieve consistency verification across multiple nodes, particularly in complex distribution networks with multiple feeders, easily leading to backfeeding or false positives for live feeders. Existing technologies also fail to consider multi-dimensional collaborative analysis of node characteristic consistency. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for detecting the reverse power transmission status for power distribution network maintenance, aiming to solve at least one of the problems in the background art.

[0006] In a first aspect, the present invention provides a method for detecting reverse power transmission status during distribution network maintenance, the method comprising:

[0007] Several detection nodes are selected in the distribution network, and each node is assigned a unique number. Each node synchronously samples voltage and current signals at a unified sampling frequency to obtain the voltage sampling sequence and current sampling sequence of the node.

[0008] The voltage sampling sequence and current sampling sequence are subjected to DC component removal and filtering to obtain preprocessed voltage and current signals;

[0009] The preprocessed signal is divided into frames according to the set frame length and frame shift. A short-time Fourier transform is performed on each frame signal to obtain the amplitude spectrum and phase spectrum of voltage and current.

[0010] Based on the preset harmonic order set, the voltage amplitude characteristics, current amplitude characteristics and phase difference characteristics corresponding to each harmonic are extracted, and the ratio characteristics of the energy of the first frequency band to the energy of the second frequency band are extracted. The phase change rate between adjacent frames is calculated to form a node feature vector, and the frequency of the first frequency band is greater than that of the second frequency band.

[0011] Each component of the node feature vector is discretized according to a preset quantization boundary set to obtain the node quantization vector, and the fingerprint digest corresponding to the node is generated by a hash algorithm through a fixed mapping relationship.

[0012] The nodes are screened for consistency based on the fingerprint digest. If the consistency is inconsistent, the nodes are compared in a fine-grained manner, and the final detection result is output based on the comparison result. The comparison result includes the maximum deviation value and the degree of consistency.

[0013] Secondly, the present invention provides a reverse power transmission prevention status detection system for distribution network maintenance, the system comprising:

[0014] The node data acquisition module is used to select several detection nodes in the distribution network, assign a unique number to each node, and synchronously sample voltage and current signals at each node according to a unified sampling frequency to obtain the voltage sampling sequence and current sampling sequence of the node.

[0015] The preprocessing module is used to remove and filter the DC component of the voltage sampling sequence and the current sampling sequence to obtain the preprocessed voltage signal and current signal.

[0016] The frame segmentation module is used to segment the preprocessed signal into frames according to the set frame length and frame shift, and perform short-time Fourier transform on each frame signal to obtain the amplitude spectrum and phase spectrum of voltage and current.

[0017] The feature extraction module is used to extract the voltage amplitude features, current amplitude features and phase difference features corresponding to each harmonic according to the preset harmonic order set, extract the ratio feature of the energy of the first frequency band to the energy of the second frequency band, and calculate the phase change rate between adjacent frames to form a node feature vector. The frequency of the first frequency band is greater than that of the second frequency band.

[0018] The fingerprint digest generation module is used to discretize each component of the node feature vector according to a preset quantization boundary set to obtain the node quantization vector, and generate the fingerprint digest corresponding to the node through a fixed mapping relationship and a hash algorithm.

[0019] The detection execution module is used to perform consistency screening on nodes based on the fingerprint digest. If the consistency is inconsistent, the nodes are compared in a fine-grained manner, and the final detection result is output based on the comparison result. The comparison result includes the maximum deviation value and the degree of consistency.

[0020] Thirdly, the present invention provides a storage medium that stores one or more programs, which, when executed by a processor, implement the above-described method for detecting the reverse power transmission status for power distribution network maintenance.

[0021] Fourthly, the present invention provides an electronic device, the electronic device comprising a memory and a processor, wherein:

[0022] The memory is used to store computer programs;

[0023] When the processor executes the computer program stored in the memory, it implements the above-mentioned method for detecting the reverse power transmission status for power distribution network maintenance.

[0024] Compared with the prior art, the present invention has the following advantages:

[0025] 1. This invention, through a multi-node synchronous sampling mechanism under unified timing at the master node, ensures that detection devices distributed across key nodes of the distribution network can acquire voltage and current signals under the same time reference. This fundamentally solves the problem of misjudgment of status caused by data asynchrony in traditional single-point detection, providing a time-consistent global data view for subsequent collaborative analysis, and achieving accurate and comprehensive perception of the energized status of the maintenance area.

[0026] 2. Multidimensional electromagnetic features, including harmonic amplitude, phase difference, phase change rate, and high-frequency / fundamental frequency energy ratio, were extracted from the synchronously sampled signals. These features together constitute a "fingerprint" vector that can finely characterize the electromagnetic state of a node. Its rich information dimensions far exceed those of traditional voltage or power direction judgments, providing a reliable data foundation for accurately identifying weak reverse power transmission signals.

[0027] 3. This invention achieves a leap from "single-point decision-making" to "swarm intelligence" by exchanging fingerprint digests and performing consistency comparisons between nodes. Even if a single node misjudges due to interference or fault, the system can still detect the anomaly by comparing its consistency with other nodes, thereby greatly improving the overall robustness and reliability of reverse power transmission status identification in complex distribution network structures (such as multiple power sources and multiple branches), and effectively avoiding local blind spots. Attached Figure Description

[0028] Figure 1 This is a flowchart of a method for detecting the reverse power transmission status for power distribution network maintenance according to an embodiment of the present invention;

[0029] Figure 2 This is a schematic diagram of the structure of a reverse power transmission prevention status detection system for power distribution network maintenance proposed in an embodiment of the present invention.

[0030] The following detailed description, in conjunction with the accompanying drawings, will further illustrate the present invention. Detailed Implementation

[0031] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Unless otherwise defined, the technical or scientific terms used herein should have the ordinary meaning understood by those skilled in the art. The terms "comprising" and similar expressions used herein mean that the element or object preceding the word covers the element or object listed after the word and its equivalents, but does not exclude other elements or objects.

[0032] like Figure 1 As shown, an embodiment of the present invention proposes a method for detecting reverse power transmission status during distribution network maintenance. The method includes steps S101 to S106, wherein:

[0033] Step S101: Select several detection nodes in the distribution network, assign a unique number to each node, and have each node synchronously sample voltage and current signals at a uniform sampling frequency to obtain the voltage sampling sequence and current sampling sequence of the node.

[0034] Step S102: Perform DC component removal and filtering on the voltage sampling sequence and current sampling sequence to obtain preprocessed voltage and current signals;

[0035] It should be noted that detection nodes are installed at key nodes of the distribution network, with node numbers i∈{1,2,...,N}.

[0036] The master reference node periodically broadcasts a time synchronization signal. Each detection node receives the synchronization signal and then determines its synchronization time based on the received time t of its local clock. r，i With the main reference time t r Calculate the time deviation δ i The node adjusts its sampling starting point according to the following formula within each sampling period:

[0037] n 0,i (p)=[pH·(1+δ i )];

[0038] Where p is the frame index, H is the number of frame shift points, and δ i Let |δ| be the clock offset of node i, if |δ| i When the allowed limit is exceeded, the node enters a pending state and requests a re-pair, n 0，i (p) is the corrected sampling starting point.

[0039] After synchronization is complete, each node proceeds according to the sampling frequency F. s Sampling is performed. The voltage sampling sequence of node i is denoted as v. u [n], the current sampling sequence is denoted as u i [n], n = 0, 1, 2, ... . The sampling period is T. s =l / F s .

[0040] To eliminate the DC offset introduced by sensor zero drift, the system calculates a moving average:

[0041]

[0042] The signal after mean correction is:

[0043]

[0044] The signal is then filtered to remove broadband noise. The filter employs a finite impulse response structure, and its output is:

[0045]

[0046] Among them, L m μ is the length of the sliding window. v，i μ u，i Let v be the estimated DC components of the voltage and current signals at node i, respectively. i [nk]、u i [nk] represents the voltage and current sample values ​​of node i at time nk, respectively. These are the corrected voltage and current sample values ​​of node i at time n, respectively. are the voltage and current sample values ​​of node i at time n after filtering, respectively, and h[l] is the coefficient of the FIR filter. These are the voltage delay signal and the current delay signal after removing the DC component, respectively. h This refers to the filter length. The filter's passband covers the power frequency and its main harmonic frequencies, while the stopband suppresses high-frequency interference.

[0047] The filtered signal undergoes amplitude normalization. Let the upper limit of the sampling system's range be A. max If the quantization bit depth is b, then the normalized signal is represented as:

[0048]

[0049] This standardization process ensures that the signal amplitudes of different nodes are within the same dimension range.

[0050] The standardized signal is segmented by frame. Let the frame length be L, the frame shift be H, and the window function be w[l], then the sampling interval of frame p is:

[0051] n = n 0，i (p)+l,l=0,1,...,L-1;

[0052] The windowed frame signal is:

[0053]

[0054] Where w[l] is the window function weight, satisfying 0≤w[l]≤1, and ∑ l w 2 [l] = 1. Each frame of signal carries a timestamp T. i，p =n 0，i (p)T s This is used to ensure the temporal consistency of subsequent feature analysis.

[0055] Step S103: The preprocessed signal is divided into frames according to the set frame length and frame shift. A short-time Fourier transform is performed on each frame signal to obtain the amplitude spectrum and phase spectrum of voltage and current.

[0056] Step S104: Extract the voltage amplitude characteristics, current amplitude characteristics and phase difference characteristics of each harmonic according to the preset harmonic order set, extract the ratio characteristics of the energy of the first frequency band and the energy of the second frequency band, calculate the phase change rate between adjacent frames, and form a node feature vector. The frequency of the first frequency band is greater than that of the second frequency band.

[0057] It should be noted that in this step, the amplitude characteristics of voltage and current are defined as the normalized ratio of the harmonic amplitude to the total amplitude of the entire spectrum, respectively; the phase difference characteristic is defined as the difference between the phase angle of voltage and current at the harmonic, and normalized to the interval (-π, π]; the phase change rate characteristic is obtained by differentiating and normalizing the phase difference of consecutive frames.

[0058] Furthermore, in some embodiments, the phase change rate is calculated according to the following formula:

[0059]

[0060] The phase difference characteristic is calculated using the following formula:

[0061]

[0062] The ratio characteristic is calculated using the following formula:

[0063]

[0064]

[0065] The voltage amplitude characteristics and current amplitude characteristics are calculated using the following formulas:

[0066]

[0067] in, Let be the phase change rate of node i in the m-th harmonic of the p-th frame, and wrap be the phase wrapping function. Let be the phase difference between the p-th frame and the (p-1)-th frame, respectively, and arg be the complex phase angle. These are the spectral values ​​of the voltage signal and the current signal at the harmonic frequency k, respectively. Let i be the ratio feature of node i in the p-th frame. The total energy of the first frequency band, The total energy of the second frequency band is given by K, where ε is a constant. H K B These are the frequency index sets for the first and second frequency bands, respectively. Let L represent the voltage amplitude characteristics and current amplitude characteristics of node i at the mth harmonic in the p-th frame, respectively, and L be the number of frequency points.

[0068] Finally, by combining all features in a fixed order, we obtain the electromagnetic fingerprint feature vector of node i in frame p:

[0069]

[0070] in, The dimension is 4M+1.

[0071] Each node outputs a set of feature vectors at each time frame, representing the electromagnetic state characteristics of that node during that time period. The feature vectors of all nodes maintain temporal and structural consistency, serving as input data for subsequent quantization processing, fingerprint digest generation, and collaborative state verification. Through this feature extraction process, the system transforms the time-domain sampled data of voltage and current signals into a multi-dimensional feature set capable of describing the electromagnetic state of the nodes, providing input data for the detection and determination of reverse power transmission status.

[0072] Step S105: Discretize each component of the node feature vector according to the preset quantization boundary set to obtain the node quantization vector, and generate the fingerprint digest corresponding to the node through a fixed mapping relationship and a hash algorithm;

[0073] It should be noted that in this step, a quantization boundary set consisting of multiple boundary values ​​is first preset for each feature component, and continuous feature values ​​are mapped to a finite integer range in order to discretize the continuous node feature vectors into node quantization vectors.

[0074] Furthermore, in some embodiments, after processing by the feature extraction unit, the electromagnetic fingerprint feature vector of node i in frame p is represented as follows:

[0075]

[0076] Where Q is the dimension of the feature vector. This represents the real value of the q-th feature component. To achieve unified transmission and summary comparison of feature information between nodes, the continuously taking feature components need to be discretized into a finite set of symbols. The system predefines the quantization boundary set for each feature component.

[0077]

[0078] in, K q This represents the number of quantization intervals for the q-th component. Based on this boundary set, the system defines a scalar quantization function Q. q (·):

[0079]

[0080] The quantization result is a sign value in integer form.

[0081] The quantization vector of a node in frame p is defined as

[0082]

[0083] A quantized vector consists of integers in a fixed order, with each component residing in the set {0, 1, ..., K}. q The value is taken from the value within the range of -1.

[0084] To reduce communication load and storage requirements, the system maps the quantization vector to a bit sequence via bit encoding. Let the q-th feature component use [log2K]... q If the binary encoding is 1, then the encoded sequence of node i is:

[0085]

[0086] Here, bin(·) represents the binary encoding function, and concat(·) represents the sequential concatenation operation. The encoded bit length is... The fingerprint digest generation unit performs a hash transformation on the encoded sequence to obtain the digest value of the node in frame p:

[0087]

[0088] in, The hash function can be a secure hash algorithm (such as SHA-256). The digest value is a fixed-length binary string, denoted as .

[0089]

[0090] Among them, L h This represents the digest bit length.

[0091] After completing quantization and summary calculation, each detection node transmits the data via a local communication link. Send to the collaborative verification unit while retaining the quantization vector. As a local backup. The communication process uses time stamp T. i,p This is to ensure that data from multiple nodes corresponds to the same time frame.

[0092] In summary, the process of node feature quantization and fingerprint digest generation includes the following steps:

[0093] (1) Receive the continuous feature vector output by the feature extraction unit;

[0094] (2) Discretize each feature component according to the preset quantization boundary set to generate an integer quantization vector;

[0095] (3) Encode the quantization results into binary and concatenate them in order;

[0096] (4) Perform a hash operation on the concatenated sequence to generate a node fingerprint digest;

[0097] (5) Send the fingerprint digest to the collaborative verification unit and save the corresponding quantization vector.

[0098] This process maps continuous signal features to discrete fingerprint identifiers, enabling different nodes to exchange state information using a unified encoding format, thus providing foundational data for subsequent collaborative verification and state consistency calculations.

[0099] Step S106: Perform consistency screening on the nodes based on the fingerprint digest. If the consistency is inconsistent, perform fine-grained comparison on the nodes and output the final detection result based on the comparison result. The comparison result includes the maximum deviation value and the degree of consistency.

[0100] In some embodiments, the steps of performing consistency screening include:

[0101] Each node exchanges the generated node fingerprint digests with each other in the local communication network, and calculates the node digest difference rate between any two node digests;

[0102] If the difference rate of the summaries of all nodes is lower than the first threshold, then the nodes are considered to be in the same state.

[0103] If at least one node has a summary difference rate greater than or equal to the first threshold, then proceed to fine-grained comparison.

[0104] Furthermore, in some embodiments, the step of performing fine-grained comparison of nodes includes:

[0105] Calculate the ratio of quantization feature differences between all pairs of nodes based on the node quantization vector, and obtain a group consistency index that characterizes the degree of consistency of the entire node group state based on the feature difference ratio.

[0106] Based on the power outage fingerprint template pre-established under the confirmed power outage state, the deviation between the current feature vector of each node and the template is calculated, and the maximum value is taken as the maximum deviation value.

[0107] More specifically, when the maximum deviation value is less than the preset template deviation threshold and the group consistency index is greater than the preset consistency threshold, it is determined to be a power outage consistency state.

[0108] When the maximum deviation value is greater than the template deviation threshold and the group consistency index is less than the consistency threshold, it is judged as a suspected reverse power transmission state.

[0109] All other cases are considered pending.

[0110] The number of times the judgment state appears in each frame is counted, and the state with the most occurrences is output as the final detection state.

[0111] Furthermore, in some embodiments, to avoid misjudgments caused by instantaneous fluctuations, the system performs majority voting smoothing in the time domain. Let the sliding window length be P, and count the occurrence frequency of each state determination result:

[0112]

[0113] The final output is the state that appears most frequently:

[0114]

[0115] The status values ​​are defined as follows: 0 indicates a consistent power outage status, 1 indicates a suspected reverse power supply status, and 2 indicates a pending status.

[0116] Final state Send to the status determination and interlocking control unit. If The interlocking control unit sends an interlocking signal; if The interlock control unit releases the interlock signal and allows maintenance; if The system maintains the current state and logs pending statuses.

[0117] Through the above steps, the collaborative verification and status determination process realizes data consistency checks between nodes, reference template deviation analysis, and time domain comprehensive determination, providing a complete logical link and executable operation process for anti-reverse power transmission status identification.

[0118] In addition, the template creation process is executed while the power outage is confirmed. Each detection node i continuously acquires the electromagnetic feature vector of frame P0.

[0119]

[0120] Where Q is the feature vector dimension and P0 is the number of template acquisition frames. The system calculates the average value of each component to form the power outage fingerprint template for node i:

[0121]

[0122] The template feature vector is represented as:

[0123]

[0124] Once the template is created, the system stores it in the template library, along with the corresponding node number and timestamp T. ref This is so that subsequent status determination calls can be made.

[0125] During long-term system operation, the template needs to be periodically updated to adapt to changes in environmental temperature, load fluctuations, and other conditions. Updates are performed under confirmed power outage conditions. Assume node i obtains a new feature vector in the k-th update cycle. The system uses a recursive averaging algorithm for template correction:

[0126]

[0127] Where α∈(0,1) are the update coefficients, and f is the template before the update. This is the updated template. The update coefficient is determined based on the rate of environmental change and is used to control the template correction magnitude. After the template update is complete, the system recalculates the template deviation threshold θ. D and consistency threshold θ C The threshold is calculated based on the sample distribution during the update period and is defined as follows:

[0128] θ D =μ D +k D σ D ,θ C =μ C -k C σ C ;

[0129] Where, μD σ D These are the sample mean and standard deviation of the template bias, respectively, μ C σ C These are the sample mean and standard deviation of the consistency index, respectively, k D k C All are empirical coefficients.

[0130] Furthermore, in some embodiments, the node summary difference rate is calculated according to the following formula:

[0131]

[0132] Calculate the characteristic difference ratio using the following formula:

[0133]

[0134] The consistency index is calculated using the following formula:

[0135]

[0136] Where C(p)∈[0,1] is the consistency index, the closer the value is to 1, the higher the consistency of features among nodes, and N is the total number of nodes. Let Q be the feature difference ratio between nodes i and j, and let Q be the feature dimension. Let be the quantization values ​​of the p-th and q-th feature components of node i in the p-th frame, respectively. L represents the node summary difference rate between node i and node j. h The length of the digest in bits, and 1() is the indicator function. , respectively, are the bits of the node fingerprint digest for nodes i and j in frame p. Let i be the node fingerprint digest of node i in frame p.

[0137] Furthermore, in some embodiments, the maximum deviation value is obtained according to the following formula:

[0138]

[0139] in, The maximum deviation value, Let i be the average normalized template bias of node i in the p-th frame. Let be the current value and template value of node i in the q-th feature component of frame p, respectively. Let T be the set of power outage fingerprint templates for node i, and T be the transpose of T.

[0140] In summary, this invention constructs a complete technical chain encompassing multi-node synchronous sampling, signal preprocessing, electromagnetic fingerprint feature extraction, quantization and fingerprint generation, collaborative verification, and interlocking linkage. First, it employs moving average and FIR filtering at each detection node to eliminate sensor zero drift and noise. Then, it extracts multi-dimensional features such as harmonic amplitude, phase difference, phase change rate, and frequency band energy ratio through short-time Fourier transform to construct the electromagnetic fingerprint. These features are then quantized and hash digests are generated for efficient cross-node comparison. Collaborative determination of the reverse power transmission status is achieved based on group consistency indicators and template deviation. A sliding time window majority voting mechanism is introduced to suppress transient misjudgments. Finally, based on the determination result, an interlocking or permission signal is automatically output, forming a closed-loop control from status perception to safe execution. This significantly improves the accuracy, reliability, and intelligence level of reverse power transmission detection during distribution network maintenance.

[0141] like Figure 2 As shown, the present invention also proposes a reverse power transmission prevention status detection system for distribution network maintenance, the system comprising:

[0142] The node data acquisition module 10 is used to select several detection nodes in the distribution network, assign a unique number to each node, and synchronously sample the voltage and current signals of each node at a unified sampling frequency to obtain the voltage sampling sequence and current sampling sequence of the node.

[0143] Preprocessing module 20 is used to perform DC component removal and filtering on the voltage sampling sequence and current sampling sequence to obtain preprocessed voltage and current signals;

[0144] The frame processing module 30 is used to perform frame processing on the preprocessed signal according to the set frame length and frame shift, and to perform short-time Fourier transform on each frame signal to obtain the amplitude spectrum and phase spectrum of voltage and current.

[0145] The feature extraction module 40 is used to extract the voltage amplitude features, current amplitude features and phase difference features corresponding to each harmonic according to the preset harmonic order set, extract the ratio feature of the energy of the first frequency band to the energy of the second frequency band, calculate the phase change rate between adjacent frames, and form a node feature vector, wherein the frequency of the first frequency band is greater than that of the second frequency band.

[0146] The fingerprint digest generation module 50 is used to discretize each component of the node feature vector according to a preset quantization boundary set to obtain the node quantization vector, and generate the fingerprint digest corresponding to the node through a fixed mapping relationship and a hash algorithm.

[0147] The detection execution module 60 is used to perform consistency screening on nodes based on the fingerprint digest. If the consistency is inconsistent, the nodes are compared in a fine-grained manner, and the final detection result is output based on the comparison result. The comparison result includes the maximum deviation value and the degree of consistency.

[0148] In another aspect, the present invention also proposes a storage medium having stored one or more programs thereon, which, when executed by a processor, implement the above-described method for detecting the reverse power transmission status for power distribution network maintenance.

[0149] In another aspect, the present invention also proposes an electronic device, including a memory and a processor, wherein the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory to implement the above-mentioned method for detecting the reverse power transmission status for power distribution network maintenance.

[0150] Those skilled in the art will understand that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can mean any means that can contain stored, communicated, propagated, or transmitted programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0151] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0152] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0153] While embodiments of the present invention have been described in detail above, it will be apparent to those skilled in the art that various modifications and variations can be made to these embodiments. However, it should be understood that such modifications and variations fall within the scope and spirit of the invention as set forth in the claims. Furthermore, the invention described herein may have other embodiments and can be implemented or carried out in various ways.

Claims

1. A method for detecting reverse power transmission status during distribution network maintenance, characterized in that, The method includes: Several detection nodes are selected in the distribution network, and each node is assigned a unique number. Each node synchronously samples voltage and current signals at a unified sampling frequency to obtain the voltage sampling sequence and current sampling sequence of the node. The voltage sampling sequence and current sampling sequence are subjected to DC component removal and filtering to obtain preprocessed voltage and current signals; The preprocessed signal is divided into frames according to the set frame length and frame shift. A short-time Fourier transform is performed on each frame signal to obtain the amplitude spectrum and phase spectrum of voltage and current. Based on the preset harmonic order set, the voltage amplitude characteristics, current amplitude characteristics and phase difference characteristics corresponding to each harmonic are extracted, and the ratio characteristics of the energy of the first frequency band to the energy of the second frequency band are extracted. The phase change rate between adjacent frames is calculated to form a node feature vector, and the frequency of the first frequency band is greater than that of the second frequency band. Each component of the node feature vector is discretized according to a preset quantization boundary set to obtain the node quantization vector, and the fingerprint digest corresponding to the node is generated by a hash algorithm through a fixed mapping relationship. The nodes are screened for consistency based on the fingerprint digest. If the consistency is inconsistent, the nodes are compared in a fine-grained manner, and the final detection result is output based on the comparison result. The comparison result includes the maximum deviation value and the degree of consistency.

2. The method for detecting reverse power transmission status for distribution network maintenance according to claim 1, characterized in that, The steps of extracting voltage amplitude features, current amplitude features, and phase difference features corresponding to each harmonic according to a preset harmonic order set, extracting the ratio feature of the energy of the first frequency band to the energy of the second frequency band, and calculating the phase change rate between adjacent frames to form a node feature vector include: The amplitude characteristics of voltage and current are defined as the normalized ratio of the harmonic amplitude to the total amplitude of the entire spectrum, respectively. The phase difference characteristic is defined as the difference between the phase angle of voltage and current at the harmonic, and normalized to the interval (-π, π]. The phase change rate characteristic is obtained by differentiating and normalizing the phase difference of consecutive frames.

3. The method for detecting reverse power transmission status for distribution network maintenance according to claim 1, characterized in that, The step of discretizing each component of the node feature vector according to a preset quantization boundary set to obtain the node quantization vector, and generating the fingerprint digest corresponding to the node through a fixed mapping relationship and a hash algorithm includes: For each feature component, a quantization boundary set consisting of multiple boundary values ​​is preset, and continuous feature values ​​are mapped to a finite integer range to discretize continuous node feature vectors into node quantization vectors.

4. The method for detecting reverse power transmission status for distribution network maintenance according to claim 1, characterized in that, The step of performing consistency screening on nodes based on the fingerprint digest includes: Each node exchanges the generated node fingerprint digests with each other in the local communication network, and calculates the node digest difference rate between any two node digests; If the difference rate of the summaries of all nodes is lower than the first threshold, then the nodes are considered to be in the same state. If at least one node has a summary difference rate greater than or equal to the first threshold, then proceed to fine-grained comparison.

5. The method for detecting reverse power transmission status for distribution network maintenance according to claim 1, characterized in that, If the consistency is inconsistent, the step of performing fine-grained comparison of the nodes includes: Calculate the ratio of quantization feature differences between all pairs of nodes based on the node quantization vector, and obtain a group consistency index that characterizes the degree of consistency of the entire node group state based on the feature difference ratio. Based on the power outage fingerprint template pre-established under the confirmed power outage state, the deviation between the current feature vector of each node and the template is calculated, and the maximum value is taken as the maximum deviation value.

6. The method for detecting reverse power transmission status for distribution network maintenance according to claim 1, characterized in that, The step of outputting the final detection result based on the comparison result includes: When the maximum deviation value is less than the preset template deviation threshold and the group consistency index is greater than the preset consistency threshold, it is determined to be a power outage consistency state. When the maximum deviation value is greater than the template deviation threshold and the group consistency index is less than the consistency threshold, it is judged as a suspected reverse power transmission state. All other cases are considered pending. The number of times the judgment state appears in each frame is counted, and the state with the most occurrences is output as the final detection state.

7. The method for detecting reverse power transmission status for distribution network maintenance according to claim 2, characterized in that, The phase change rate is calculated using the following formula: The phase difference characteristic is calculated using the following formula: The ratio characteristic is calculated using the following formula: The voltage amplitude characteristics and current amplitude characteristics are calculated using the following formulas: in, Let be the phase change rate of node i in the m-th harmonic of the p-th frame, and wrap be the phase wrapping function. Let be the phase difference between the p-th frame and the (p-1)-th frame, respectively, and arg be the complex phase angle. These are the spectral values ​​of the voltage signal and the current signal at the harmonic frequency k, respectively. Let i be the ratio feature of node i in the p-th frame. The total energy of the first frequency band, The total energy of the second frequency band is given by K, where ε is a constant. H K B These are the frequency index sets for the first and second frequency bands, respectively. Let L represent the voltage amplitude characteristics and current amplitude characteristics of node i at the mth harmonic in the p-th frame, respectively, and L be the number of frequency points.

8. The method for detecting reverse power transmission status for distribution network maintenance according to claim 4, characterized in that, The node summary difference rate is calculated using the following formula: Calculate the characteristic difference ratio using the following formula: The consistency index is calculated using the following formula: Among them, C (p) ∈[0,1] is the consistency index; the closer the value is to 1, the higher the consistency of features among nodes. N is the total number of nodes. Let Q be the feature difference ratio between nodes i and j, and let Q be the feature dimension. Let be the quantization values ​​of the p-th and q-th feature components of node i in the p-th frame, respectively. L represents the node summary difference rate between node i and node j. h The length of the digest in bits, and 1() is the indicator function. , respectively, are the bits of the node fingerprint digest for nodes i and j in frame p. Let i be the node fingerprint digest of node i in frame p.

9. The method for detecting reverse power transmission status for distribution network maintenance according to claim 1, characterized in that, The maximum deviation value can be obtained using the following formula: in, The maximum deviation value, Let i be the average normalized template bias of node i in the p-th frame. Let be the current value and template value of node i in the q-th feature component of frame p, respectively. Let T be the set of power outage fingerprint templates for node i, and T be the transpose of T.

10. A reverse power transmission prevention status detection system for power distribution network maintenance, characterized in that, The system includes: The node data acquisition module is used to select several detection nodes in the distribution network, assign a unique number to each node, and synchronously sample voltage and current signals at each node according to a unified sampling frequency to obtain the voltage sampling sequence and current sampling sequence of the node. The preprocessing module is used to remove and filter the DC component of the voltage sampling sequence and the current sampling sequence to obtain the preprocessed voltage signal and current signal. The frame segmentation module is used to segment the preprocessed signal into frames according to the set frame length and frame shift, and perform short-time Fourier transform on each frame signal to obtain the amplitude spectrum and phase spectrum of voltage and current. The feature extraction module is used to extract the voltage amplitude features, current amplitude features and phase difference features corresponding to each harmonic according to the preset harmonic order set, extract the ratio feature of the energy of the first frequency band to the energy of the second frequency band, and calculate the phase change rate between adjacent frames to form a node feature vector. The frequency of the first frequency band is greater than that of the second frequency band. The fingerprint digest generation module is used to discretize each component of the node feature vector according to a preset quantization boundary set to obtain the node quantization vector, and generate the fingerprint digest corresponding to the node through a fixed mapping relationship and a hash algorithm. The detection execution module is used to perform consistency screening on nodes based on the fingerprint digest. If the consistency is inconsistent, the nodes are compared in a fine-grained manner, and the final detection result is output based on the comparison result. The comparison result includes the maximum deviation value and the degree of consistency.

Images